10
Biochemical Pharmacology, Vol. 21, pp. 1217-1226. Pergamon Press, 1972. Printed in Great Britain. EFFECT OF Ag-TETRAHYDROCANNABINOL ON MITOCHONDRIAL PROCESSES* JOAN MUNROE MAHONEY and ROBERT A. H-s? Department of Biochemistry, Indiana University School of Medicine, Indianapolis, Ind. 46202, U.S.A. (Received 8 October 1971; accepted 10 December 1971) Abstract-The major psychoactive component of marijuana, Ag-tetrahydrocannabinol (THC), strongly at&ted rat liver mitochondria in vitro. At concentrations of 154 nmoles/mg of mitochondrial protein, THC uncoupled state IV respiration and decreased respiratory control and ADP/O ratios. Energy-linked changes in fluorescence of 8- aniline-1-naphthalene sulfonate were prevented or reversed by THC. THC also pro- duced large amplitude swelling of mitochondria and release of matrix enzymes. These effects were greatly potentiated by Mg =+. Likewise, flocculation of mixed phospho- lipid micelles by MgZ + was potentiated greatly by low concentrations of THC. Studies with micelles prepared from purified phospholipids suggest that THC may specitically destabilize the cardiolipin in mixed micelles. OF THE many cannabinoids found in marijuana and its active extracts, the major psychoactive compound is Ag-tetrahydrocannabinol (THCS).l Its structure is shown in Fig. 1. Recent isolation and purification of this material2 permits direct testing of its influence on subcellular organelles. Because of the great insolubility in water and the polycyclic structure of this compound, it was predicted that THC would have rather strong effects upon membrane-dependent processes. The hydrophobic portion ‘CH3 A9-Tetrahydrocarmabinol FIG. 1. Structure of Ag-tetrahydrocamrabinol (‘D-X). * Supported in part by grants from the N.I.H. (HEO4219-12, HE06308, and MH 15864-01). t Established Investigator of American Heart Association. $ Abbreviations used: THC, Ag-tetrahydrocannabinol (atoms numbered according to the dibenxo- pyran system); RLM, rat liver mitochondria; NAD+, nicotinamide adenine dinucleotide; ADP, adenosine diphosphate; ATP, adenosine triphosphate; P,, inorganic phosphate; RCR, respiratory control ratio; ADP/O, nanomoles of ADP added per nanoatoms of extra O2 consumed; ANS, 8-ani- lino-l-naphthalene sulfonate; MDH, malate dehydrogenase. B.P. 21/9-A 1217

Effect of Δ9-tetrahydrocannabinol on mitochondrial processes

Embed Size (px)

Citation preview

Page 1: Effect of Δ9-tetrahydrocannabinol on mitochondrial processes

Biochemical Pharmacology, Vol. 21, pp. 1217-1226. Pergamon Press, 1972. Printed in Great Britain.

EFFECT OF Ag-TETRAHYDROCANNABINOL ON MITOCHONDRIAL PROCESSES*

JOAN MUNROE MAHONEY and ROBERT A. H-s?

Department of Biochemistry, Indiana University School of Medicine, Indianapolis, Ind. 46202, U.S.A.

(Received 8 October 1971; accepted 10 December 1971)

Abstract-The major psychoactive component of marijuana, Ag-tetrahydrocannabinol (THC), strongly at&ted rat liver mitochondria in vitro. At concentrations of 154 nmoles/mg of mitochondrial protein, THC uncoupled state IV respiration and decreased respiratory control and ADP/O ratios. Energy-linked changes in fluorescence of 8- aniline-1-naphthalene sulfonate were prevented or reversed by THC. THC also pro- duced large amplitude swelling of mitochondria and release of matrix enzymes. These effects were greatly potentiated by Mg =+. Likewise, flocculation of mixed phospho- lipid micelles by MgZ + was potentiated greatly by low concentrations of THC. Studies with micelles prepared from purified phospholipids suggest that THC may specitically destabilize the cardiolipin in mixed micelles.

OF THE many cannabinoids found in marijuana and its active extracts, the major psychoactive compound is Ag-tetrahydrocannabinol (THCS).l Its structure is shown in Fig. 1. Recent isolation and purification of this material2 permits direct testing of its influence on subcellular organelles. Because of the great insolubility in water and the polycyclic structure of this compound, it was predicted that THC would have rather strong effects upon membrane-dependent processes. The hydrophobic portion

‘CH3

A9-Tetrahydrocarmabinol

FIG. 1. Structure of Ag-tetrahydrocamrabinol (‘D-X).

* Supported in part by grants from the N.I.H. (HEO4219-12, HE06308, and MH 15864-01). t Established Investigator of American Heart Association. $ Abbreviations used: THC, Ag-tetrahydrocannabinol (atoms numbered according to the dibenxo-

pyran system); RLM, rat liver mitochondria; NAD+, nicotinamide adenine dinucleotide; ADP, adenosine diphosphate; ATP, adenosine triphosphate; P,, inorganic phosphate; RCR, respiratory control ratio; ADP/O, nanomoles of ADP added per nanoatoms of extra O2 consumed; ANS, 8-ani- lino-l-naphthalene sulfonate; MDH, malate dehydrogenase.

B.P. 21/9-A 1217

Page 2: Effect of Δ9-tetrahydrocannabinol on mitochondrial processes

1218 JOAN MUNROE MAHONEY and ROBERT A. HARRLS

of the molecule would be expected to interact with hydrophobic regions in the mem- brane and perhaps perturb the functioning of the mitochondrial membrane system. This study describes the effects of THC upon respiration, swelhng, and energy-linked changes in fluorescence of 8-aniline-l-naphthalene sulfonate (ANS) with rat liver mitochondria (RLM).

MATERIALS AND METHODS

Rat liver mitochondria (RLM) were prepared as described by Johnson and Lardy. 3 Oxygen ~nsumption was measured with an oxygen electrode as described previouslyP The incubation medium (5 ml) was O-225 M in sucrose, 10 mM in P, (pH 7*4), 5 mM in MgCI,, 20 mM in KCl, and 20 mM in triethanolamine. THC was a gift from Dr. Robert B. Forney and was isolated from marijuana and purified to greater than 99 per cent purity by the procedure described by Turk et aL2 Since the light brown oil dissolved readily in 95 % ethanol, small volumes (< O-03 ml) of alcoholic solutions of THC were added to the mitochondrial incubation medium described above. Swelling of the mitochondria was measured by changes in the optical density at 540 nm in 3 ml of incubation medium (as described above) cont~ning 05-0~8 mg of RLM protein.

For studies of the release of mitochondrial malate dehydrogenase, MgCl, was omitted from the incubation medium. To 3 ml of Mg-free medium were added 1 pg of antimycin and 0.5 mg of RLM protein. After swelling had been monitored at 520 nm for a few min at 30”, the contents of the cuvette were transferred to 1*5-ml plastic tubes and centrifuged in an Eppendorf centrifuge for 5 min. The supernatant was removed carefully with a Pasteur pipette and stored on ice for assay of malate dehydro- genase. This assay was done at 38” in a volume of 1 ml of reaction medium conning 90 pmoles of glycine-NaOH buffer (PH 9~9)~ 100 pmoles of malate, 2-5 pmoles of neutralized NAD+, and 1 pg of antimycin. The production of NADH was followed at 340 nm. To permit determination of the total malate dehydrogenase activity present in mitochondria before incubation, the RLM were disrupted with deoxycholate.

Fluorescence was measured with an Aminco Bowman Spectrophotofluorometer with a front-face attachment. Excitation was at 365 nm and fluorescence was followed at 480 nm. The reaction medium was 0.225 M in sucrose, 20 mM in KCl, 20 mM in triethanolamine, and also contained, in a total volume of 3 ml, 0.3 rmoles of ANS, 5 pg of rotenone, and 3 mg of RLM protein.

Phospholipids were purified from heavy beef heart mitochondria by the method of Rouser and Fleischer.5 Purified cardiolipin was purchased from Sylvanna Chemical Company, Orange, N.J. Micelles were prepared by pipetting the desired quantity of phospholipid in chloroform into a small beaker (10-30 ml), evaporating the chloro- form with a stream of nitrogen, and adding a medium O-25 M in sucrose, and 10 mM in tris Cl, pH 7-4. The mixture was sonicated with a Bronwill Biosonik Model BP1 sonicator on ice for three 1-min bursts alternating with l-mm cooling intervals. Triton micelles were prepared from Triton X-100 (Calbiochem, B grade) by the same procedure. The suspension was transferred to a test tube and allowed to equilibrate in a 30” water bath for 1 hr. Micelles of mixed mitochondrial phospholipids or Triton X-100 were made at a concentration of 10 mg (dry weight)/mI, and 0-I ml was added to the incubation cuvette to give a final concentration of 1 mg in 3 ml. Phosphatidyl- choline, cardio~pin, and a mixture of the two were sonicated at a lipid concentration

Page 3: Effect of Δ9-tetrahydrocannabinol on mitochondrial processes

THC and mitochondria 1219

of 4 mg in 12 ml and used without dilution. Flocculation was monitored by following the optical density at 520 nm.

RESULTS

THC increased state IV respiration at concentrations of 15-60 nmoles/mg of RLM protein with succinate and NAD-linked substrates. The respiratory control ratio (state III/state IV) was significantly decreased in all cases with these levels of THC, whereas the ADP/O ratio was decreased at concentrations greater than 30 nmoles/mg of protein. At still higher concentrations, THC inhibited respiration with NAD-linked substrates. Addition of similar volumes of ethanol to control incubations produced none of these changes. Typical results are shown in Fig. 2.

AD/ RCR=P6

\

ADP/O = 1.0

FIG. 2. Effect of THC on oxygen consumption by RLM. RLM (2 mg protein/ml) were incubated in 5 ml of the medium described in the text. Further additions were: (A) 10 pmoles of glutamate and 2.5 pmoles of malate; 895 nmoles ADP/addition and 36 nmoles THC/mg protein were added where indicated on the trace. 03) As in A. exceot 72 nmoles THC/mg nrotein were used. (Cl 10 timoles of succinate and 5 pg of rotenone; 537 nmoles of ADPjaddition-and 36 nmoles of TIk/mg protein were added where indicated on the trace. The respiratory control ratio (RCR) and ADP/O ratios are

indicated on the trace for each addition of ADP.

Under conditions of the assay for respiration, the lower concentrations of THC that caused uncoupling did not appear to cause swelling of the RLM in media containing 5 mM MgCl,. At the upper range of uncoupling concentrations and at still higher concentrations (> 60 nmoleslmg protein), THC led to rapid and extensive large amplitude swelling in the presence of Mg 2+. Traces of optical density changes which demonstrate the magnitude of the swelling are presented in Fig. 3. Failure of rotenone, antimycin, and oligomycin to inhibit swelling induced by THC suggests that energy is not required for this action. The presence of an oxidizable substrate was not required.

Page 4: Effect of Δ9-tetrahydrocannabinol on mitochondrial processes

1220 JOAN MUNROE MAHONEY and ROBERT A. HARRIS

FIG. 3. THC-induced swelling of RLM. The incubation medium (3 ml) contained 10 pmoles of gluta- mate, 2.5 pmoles of malate, 1 pmole of ADP, and O-8 mg of RLM protein. Curve A, control (no THC). Curve B, 30 nmoles THCfmg protein were added at the break in the curve. Curve C, 60 nmoles

THC/mg protein were added.

The swelling of mitochondria induced by THC was greatly facilitated by Mg*“, as seen in Fig. 4. The large amplitude swelling in the presence of THC and Mg*+ was accompanied by release of most of the mitochondrial malic dehydrogenase, a marker

enzyme for the matrix space.

I f AOD = 0.30

%QH Released

1.8 2.6

7.3

86.6 EIG. 4. Effect of THC and Mg2 + on swelling and release of malate dehy~o~nase (MDH). The incu- bation medium (3 ml) contained 1 pg antimycin and O-5 mg RLM protein. (A) Control without MgC12. (B) Control with 5 mM MgCl,. (C) No MgC12. THC, 60 nmoles/mg protein, was added at the break in the curve. (D) In the presence of 5 mM MgC12, THC, 60 nmoIes/mg protein, was added at the break in the curve. The release of MDH into the supernatant medium is indicated for each

samole.

Page 5: Effect of Δ9-tetrahydrocannabinol on mitochondrial processes

THC and mitochondria 1221

Changes in fluorescence of ANS have been shown to correlate well with the energy state of the mitochondrion.6 The effects of THC on the fluorescence of ANS in the presence of rat liver mitochondria are shown in Fig. 5. In the presence of Mg2+, 67 nmoles of THC/mg of protein increased the fluorescence of ANS in mitochondrial

A C HATP

FIG. 5. Effect of THC and Mg2+ on energy-linked changes in ANS fluorescence. To the medium described in the text were added as indicated on the trace: 5 pmoles succinate (WCC), 0.6 pg antimycin (ant), 5 pmoles ATP, 1.5 pg oligomycin (oligo), and 200 nmoles THC (67 nmoles/mg protein). Upward deflection represents an increase in fluorescence as indicated by the arrow. (A) Control without MgClz. (ES) Effect of THC without MgCl,. (C) Control with 5 mM Mg&. (D) Effect of THC with

5 mM MgC12.

suspensions oxidizing succinate and prevented decreases caused by the addition of ATP. This concentration of THC was not effective in the absence of Mg2 +. Other studies also showed that THC at 67 nmoles/mg protein in the presence of MgZf (but not in the absence) would prevent the change induced by succinate oxidation and would reverse the change caused by ATP. Higher concentrations of THC reversed and prevented energy-linked changes with or without Mg2+. Either the higher concen- trations of THC caused more deleterious changes or the amount of Mg2+ normally associated with the mitochondria was sufficient for the action of these higher doses.

Studies of the effect of THC on isolated phospholipids of mitochondria are shown in Fig. 6. THC had little effect on micelles of mixed mitochondrial phospholipids in the absence of Mg 2+, but subsequent addition of Mg 2+ led to rapid flocculation of the micelles. Other studies showed that addition of this concentration of Mg2+ first did not lead to flocculation until THC was added. This was not a general behavior of all material in micellar dispersement, as shown in Fig. 7. THC produced little effect

Page 6: Effect of Δ9-tetrahydrocannabinol on mitochondrial processes

1222 JOAN MUNROE MAHONEY and ROBERT A. HARRIS

t-2min-4

T

AOD = 0.01

1.

THC +

FIG. 6. Flocculation of mixed mitochondrial phospholipids by THC. The optical density of a sus- pension of mixed mitochondrial phospholipids, prepared as described in the text and containing 1 mg phospholipid (dry weight) in 1 ml, was followed for 2 min. To one sample 120 nmoles THC in 0.03 ml 95 % ethanol was added, and 0.03 ml 95 % ethanol was added to the other. The optical density was followed for about 2 more minutes. Then MgC12 was added to each sample to a fmal concentra- tion of 5 mM. An increase in O.D. represents an increase in turbidity of the suspension because of

flocculation of the micelles.

on micelles of the water soluble detergent, Triton X-100, in the presence or absence of Mg’+ (Fig. 7A). Figure 7B shows that THC, with or without Mg’+, had no effect on micelles of purified phosphatidylcholine prepared from the mitochondrial phos- pholipids used in Fig. 6. On the other hand, micelles of cardiolipin were flocculated immediately by Mg2+ whether THC was present or not as seen in Fig. 7C. This was a very rapid and extensive change, the full scale deflection in Fig. 7C representing ten times the deflection of Figs. 6-7B. After examination of two cases of purified phospho- lipids where THC and Mg 2+ had no effect on the stability of their micelles and one case where THC was not needed for a large effect of Mg’+, it was found that mixed micelles of phosphatidylcholine plus cardiolipin paralleled the results obtained with micelles of mixed mitochondrial phospholipids. An example of these results is seen in Fig. 8. Addition of THC without Mg2+ (or Mg2+ without THC) had little effect on mixed micelles of phosphatidylcholinecardiolipin 2 : 1 (w/w), but subsequent addition of Mg2+ caused extensive flocculation of THC-sensitized micelles.

DISCUSSION

These studies demonstrate that THC is a fairly active reagent with regard to alter- ing the structure and function of the rat liver mitochondrial system. In this respect, THC is not as powerful at uncoupling mitochondria as a number of other lipophilic compounds. Under the conditions of the assays reported here, respiratory control was

Page 7: Effect of Δ9-tetrahydrocannabinol on mitochondrial processes

THC and mitochondria 1223

c

f AOD = 0’10

_I_

FIG. 7. Effect of THC on micelles of other substances. The clear micelles were prepared as described in the text, and additions were as in Fig. 6. ,(A) The solution contained 1 mg Triton X-100 in 3 ml. (B) The solution contained 1 mg purified phosphatidylcholine in 3 ml. (C) The solution contained

1 mg purified cardiolipin in 3 ml.

significantly decreased by THC at concentrations of 3 x 10m5 M (15 nmoles/mg of mitochondrial protein), whereas a decrease in the ADP/O ratio was observed only with concentrations in excess of 6 x IO-§ M (30 nmoles/mg of mitochondrial protein). A number of “uncouplers” of mitochondrial processes, e.g. the derivatives of carbonyl cyanide phenylhydrazone, uncouple at much lower concentrations than THC. For example, carbonyl cyanide p-trifluoromethoxyphenylhydrazone exhibits uncoupling effects at concentrations as low as 10V8 M’. Since THC appears quali- tatively to be more hydrophobic than those compounds which have greater potency as uncouplers, the metabolites of THC which are more polar may show greater activity upon mitochondrial processes. THC has been shown to be metabolized to a number of more polar compounds,s including 11-hydroxy THC. Although a minor metabolite, the latter is active psychomimeticallys*9 and has not yet been tested upon mitochondrial processes.

This study indicates that THC induces swelling and lysis of mitochondria in the presence of Mg 2+. Byington et aLlo and Smoly et al. l l also found that other relatively hydrophobic compounds caused swelling with Mg2 + . In agreement with this study, they found that the swelling was accompanied by release of the matrix enzymes, malate dehydrogenase, isocitrate dehydrogenase and a-ketoglutarate dehydrogenase. Di- ethylstilbestrol and steroid hormones such as progesterone and 4-chloro-17-methyl-

Page 8: Effect of Δ9-tetrahydrocannabinol on mitochondrial processes

1224 JOAN MUNROE MAHONEY and ROBERT A. HARRIS

AOD= O-01

1

// EtOH

FIG. 8. Effect of THC on mixed micelles of phosphatidylcholine and cardiolipin. A micellar suspension containing O-67 mg purified phosphatidylcholine and 0.33 mg purified cardiolipin in 3 ml was

prepared as described in the text. Additions were as in Fig. 6.

19nortestosterone had similar fragmenting capability. Lipophilic ionophoretic anti- biotics such as valinomycin and gramicidin D would produce the same effects at high concentrations. Compounds too polar, e.g. hydrocorticosterone, or too apolar to be very soluble in ethanol, e.g. cholesterol, were shown to be inactive. Likewise, agents with significant water solubility (water-soluble polypeptide antibiotics such as baci- tracin, colistin, ristocetin, polymixin, and vancomycin) were without activity over a wide range of concentrations.10 THC is typical of the class of active hydrophobic materials.

Many studies have established that cations interact with phospholipids (e.g. Shah and Schulman). I2 Many years ago, H. G. Bungenberg de Jong13 studied the effects of ions on egg lecithin suspensions. He found that clear sols of egg lecithin would not precipitate, or would precipitate only slowly, upon the addition of a number of monovalent or divalent cations. He also found that the sols could be sensitized to precipitation by adding a number of organic non-electrolytes such as cholesterol or triolein.13 His results are somewhat similar to the ones reported in this study, where THC sensitized phosphatidylcholine-cardiolipin micelles to flocculation by Mg2 + .

In a micelle of phospholipid, the fatty tails are oriented inward, away from the water, where they stabilize the micelle by hydrophobic interactions. The surface of the micelle

Page 9: Effect of Δ9-tetrahydrocannabinol on mitochondrial processes

THC and ~tochondria 1225

consists of the partially, or fully ionized groups of the polar portion of the molecules, surrounded by counterions from the suspending medium and their water of hydra- tion.14 If the micelles carry an effective net charge, mutual repulsion of the particles helps prevent them from coalescing. hydration of the ion&d groups on the surface of the micelle also stabilizes the structure. Ions with a large capability for hydration, like

Mg’+, may remove some of the hydration water that stabilized the micelle, and thus lead to cross-linking of micelles, coagulation, and precipitation.14 Mg*+, in counter- balancing two negative charges on one cardiolipin molecule, may perturb the con- formation of the molecule so that the micelle is destabilized. Charge repulsion must play such an impo~ant role in stabi~~ng ~rdiolipin micelles that Mg2 + will flocculate them even without the added destabilizing influence of THC. In mixed phospholipid micelles, on the other hand, the stabilizing effects of both water of hydration and charge repulsion are balanced so that the added destabilizing effect of THC is needed for precipitation.

Chari-Bitron and Bino15 have studied the effect of A8-tetrahydrocannabinol on the ATPase activity of rat liver mitochond~a. They also noticed a great stimulation by Mg2+ of the effects of As-tetrahydrocannabinol. Their conclusion that As-tetrahydro- cannabinol must cause some disorganization of the mitochondria is corroborated by this study. Although they used the A8 isomer and we used the Ag, the effects are quite similar and both are psychoactive. l THC reacts with the membrane in such a way as to cause the mitochondrion to swell and fragment to the extent that matrix enzymes are released. The mechanism is not completely clear but phospholipids and parti- cularly cardiolipin are implicated by our studies. The mitochondrial inner membrane bears a net negative charge l6 because of the preponderance of negatively charged phospholipids, and probably is stabilized by appreciable water of hydration. THC may destabilize the membrane so that the components are more dependent on mutual repulsion between charged groups for stability. Adsorption of lvIg2+ would neutralise the negative charges and destabilize the structure as with phospholipid micelies. Additional studies with this type of reagent may help further to define the forces involved in stabilizing biological membranes.

Whether the effects of THC noted in this study are related in any way to the action of THC irt vim is unknown. The Ievels which were necessary to show appreciable effects upon mitochond~al processes were rather high. In addition, extensive damage to the membrane (probably irreversible) was observed at the higher concentrations used. Perhaps the most significant observation of this study with respect to the mechanism of action of THC in vivo is the demonstration of the relationship between THC and divalent cations. The latter are known to be important in proper nerve function and this study suggests their interaction with the membrane may be greatly influenced by THC.

Acknowledgemenl thank Mrs. Robert Sisk and Miss Pat Jenkins for technical assistance and Drs. David M. Gibson and Robert B. Foney for their interest and support in this research.

REFERENCES 1. R. MESON, A. SH~NI, H. EDERY and Y. GRUNFELD, Science, N. Y. l69,611 (1970). 2. R. F. TURK, R. B. FORNEY, L. B. KING and S. RAMACHANDRAN, J. forens. Sci. 14, 385 (1969). 3. D. JOHNSON and H. LARLIY, in Methods in Enzymology (Eds. R. W. ESTABROOK and M. E.

PULLMAN), Vol. X, p, 94. Academic Press, New York (1967).

Page 10: Effect of Δ9-tetrahydrocannabinol on mitochondrial processes

1226 JOAN MUNROE MAHONEY and ROBERT A. HARRIS

4. R. A. HARRIS, J. MIJNROE, B. FARMER, K. C. KIM and P. JENKINS, Archs Biochem. Biophys. 142, 435 (1971).

5. G. ROUSER and S. FLEISCHER, in Methods in Enzymology (Eds. R. W. ESTAEIR~~K and M. E. PULLMAN), Vol. X, p. 385. Academic Press, New York (1967).

6. A. AZZI, Biochem. biophys. Res. Commun. 37,254 (1969). 7. P. G. HEyTL.J?R and W. W. PRITCHARD, Biochem. biophys. Res. Commun. 7,272 (1962). 8. L. LEMBERGER, S. D. SILBERSTEIN, 3. AXELROD and I. J. KOPIN, Science, N. Y. 170,132O (1970). 9. R. L. FOLTZ, A. F. FENTIMAN, JR., E. G. LEIGHTY, J. L. WALTER, H. R. DRE~ES, W. E. SCHWARTZ,

T. F. PAGE, JR. and E. B. TRUITT, JR., Science, N. Y. 168,844 (1970). 10. K. H. BYINGTON, J. M. SMOLY, A. V. MOREY and D. E. GREEN, Archs Biochem. Biophys. 128,

762 (1968). 11. J. M. SMOLY, K. H. BYINGTON, W. C. TAN and D. E. GREEN, Archs Biochem. Biophys. 128,774

(1968). 12. D. 0. SHAH and J. H. SCH~LMAN, J. Lipid Res. 6, 341 (1965). 13. H. G. BUNGENBERG DE JONG, in Colloid Science (Ed. H. R. KRUYT), Vol II, p. 276. Elsevier, New

York (1949). 14. K. J. MYSEU, Introduction to Colloid Chemistry, pp. 299-373. Interscience, New York (1959). 15. A. CHARI-BITRON and T. BINO, Biochem. Pharmac. 20,473 (1971). 16. G. A. BL~NDIN and D. E. GREEN, J. Bioenergetics 1, 193 (1970).